Memory, the first thing to fray in Alzheimer’s, flickers back on in animals when a humble cellular molecule is topped up. In an international study published November 7, 2025 in Science Advances, researchers report that boosting NAD+ restores learning and memory in worm and mouse models by correcting RNA splicing mistakes through a protein called EVA1C.
The team, led by Evandro Fei Fang at the University of Oslo with collaborators in China, Portugal, the UK, Japan, Greece, and Spain, ties a long suspected brain-protective metabolite to one of Alzheimer’s quieter culprits, faulty RNA splicing. In Alzheimer’s and aging, genes often produce the wrong protein isoforms, a mis-edit that can derail synapses and metabolism. By raising NAD+ using precursors such as nicotinamide riboside (NR) or nicotinamide mononucleotide (NMN), the scientists show those edits can be steered back toward healthy patterns, easing tau-driven damage and rescuing behavior.
Across species, the story is consistent. In roundworms engineered to express human tau, NAD+ normalized neuron-specific splicing and improved memory-like chemotaxis behavior. In mice carrying tau pathology, weeks of NAD+ precursor treatment restored performance on a standard novel object test without altering locomotion or body weight. The mechanistic linchpin is EVA1C, a splicing-related protein whose levels and isoforms shift in disease and after NAD+ augmentation. When the researchers knocked down EVA1C, the benefits of NAD+ vanished.
“Notably, we found when the EVA1C gene was knocked down, these benefits were lost, confirming that EVA1C is essential for NAD+-mediated neuroprotection”.
What makes the finding feel concrete is the microscopy. In hippocampal slices, EVA1C signals overlap with HSP70, a heat-shock chaperone, under blue DAPI nuclear stain and green HSP70 channels, while colocalization with BAG1 drops after NR treatment. That shift hints at a chaperone network realignment downstream of corrected splicing. The group used AlphaFold-guided modeling to predict how specific EVA1C isoforms prefer HSP70 over BAG1, then backed that up with co-immunoprecipitation and quantitative image analysis.
The transcriptomics add scale. Differential alternative splicing events, especially exon skipping and splice-site changes, were mapped across hundreds of genes tied to axon guidance, mitochondrial trafficking, autophagy, and oxygen metabolism. In tauopathy mice, many of these pathways were off-kilter. After NAD+ augmentation, splicing profiles moved toward wild-type, and compensatory overexpression of spliceosome components settled down, a sign that the system no longer needed emergency workarounds.
From Molecule To Mechanism
NAD+ is best known for shuttling electrons in energy metabolism, but it also serves as a currency for enzymes that regulate chromatin and RNA processing. Here, the authors connect that metabolic coin to a precise editing outcome: EVA1C-dependent splice corrections that ripple across synaptic and proteostasis genes. In a vivid laboratory scene, tau-burdened neurons glow with altered patterns under the microscope; after NAD+ support, the red EVA1C puncta line up more often with green HSP70, suggesting a restored quality-control assembly line.
Human relevance appears in databases and tissue. EVA1C transcripts are elevated in Alzheimer’s cohorts, with region-specific patterns in parahippocampal and temporal cortex, while neuronal EVA1C protein drops across Braak stages in hippocampus and entorhinal cortex. That RNA–protein mismatch, common in neurodegeneration, aligns with the study’s central claim that splicing control, not just expression level, matters.
The work does not claim a cure, and the authors note caveats: short treatment windows, isoform-specific antibodies still in development, and the need to test induced neurons and organoids. Still, the convergence across species and modalities makes the mechanistic picture unusually crisp.
“We propose that maintaining NAD+ levels could help preserve neuronal identity and delay cognitive decline, paving the way for combination treatments to enhance RNA splicing”.
Therapeutic Next Steps
Clinically, NAD+ precursors already under study for safety and cognition may benefit from this mechanistic rationale. The data argue for trials that stratify by splicing biomarkers and track EVA1C-linked networks, not only standard cognitive endpoints. Combination approaches are also on the table, pairing NAD+ augmentation with splice-switching oligonucleotides aimed at tau and synaptic genes, or with chaperone modulators that reinforce the HSP70 axis highlighted here.
The broader takeaway lands with a gentle thud: in Alzheimer’s, metabolism and message editing are entwined. Fix the cell’s energy bookkeeping, and you may also fix how genes are read, trimmed, and assembled into the proteins that make thought possible. It is a sober hope, not a headline promise, but it is grounded in data, cross-validated, and testable in people.
Science Advances: 10.1126/sciadv.ady9811
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.